Assessing deep learning models for multi-class upper endoscopic disease segmentation: A comprehensive comparative study
Hierarchical architectures like Swin-UMamba and SegFormer show promise for UGI disease segmentation, reducing missed diagnoses and improving workflows, but robust clinical validation is crucial for real-world deployment.
Authors
- Stephen T.C. Wong (ORCID: https://orcid.org/0000-0001-9188-6502)
- Ye-Ying Qin
- In Neng Chan (ORCID: https://orcid.org/0000-0002-7015-4510)
- Baoliang Zhao (ORCID: https://orcid.org/0000-0002-8157-5951)
- Chi Hong Wong (ORCID: https://orcid.org/0000-0002-3238-3793)
- Pak Kin Wong (ORCID: https://orcid.org/0000-0002-7623-6904)
- Ying Hu (ORCID: https://orcid.org/0000-0002-9711-6047)
- Hon Ho Yu (ORCID: https://orcid.org/0000-0002-9580-345X)
- In Weng Chan
- Zheng Li (ORCID: https://orcid.org/0000-0003-0033-5938)
- Shan Gao (ORCID: https://orcid.org/0000-0003-1950-0729)
- Liang Yao (ORCID: https://orcid.org/0000-0003-4068-3136)
- Tao Yan (ORCID: https://orcid.org/0000-0002-8929-015X)
- Chon In Chan (ORCID: https://orcid.org/0009-0009-0999-5028)
- Yanyan Hu (ORCID: https://orcid.org/0009-0009-3809-9505)
- Ieng Hou Lam
Institutions
- Macau University of Science and Technology (MO)
- University of Macau (MO)
- Hubei University of Arts and Science (CN)
- Shenzhen Institutes of Advanced Technology (CN)
Publication Details
- Journal
- World Journal of Gastroenterology
- Published
- 2025-11-06
- DOI
- https://doi.org/10.3748/wjg.v31.i41.111184
- Citations
- 1
- Primary Topic
- Colorectal Cancer Screening and Detection
- Type
- article
- Field-Weighted Citation Impact
- 1.05